Complex-valued radial basis function network, Part II: Application to digital communications channel equalisation

Chen, Sheng; McLaughlin, Stephen; Mulgrew, Bernard

doi:10.1016/0165-1684(94)90206-2

Cited by 129 publications

(59 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In conclusion, the performance of the RBF DFE has been documented in the context of fixed-mode modulation schemes in [6] and [7]. However, there is little information concerning its potential in either BbB AQAM or in FEC-coded scenarios, despite the advantageous interactions of RBF-aided DFE BbB AQAM in conjunction with turbo FEC, which were demonstrated in our preliminary work [8].…”

Section: Discussionmentioning

confidence: 97%

“…Wong et al [5] extended these contributions to dispersive wideband channels with the aid of a burst-by-burst (BbB)-adaptive Kalman filtered AQAM scheme. Radial basis function (RBF)-assisted decision feedback equalization (DFE) has been documented in the context of fixed-mode modulation schemes [6], [7]. However, there is little information concerning its potential in either BbB AQAM or in forward error correction (FEC)-coded scenarios, despite the advantageous interactions of RBF-aided DFE BbB AQAM in conjunction with turbo FEC, which were demonstrated in our preliminary work [8].…”

Section: Introductionmentioning

confidence: 99%

“…The equalizer's input is the -dimensional channel output vector at signaling instant . The RBF network has a similar structure to the optimal Bayesian equalizer [16] and thus it was proposed for equalization in [6], [16], [17]. The output of the RBF equalizer based on the optimal Bayesian decision function-where the RBF centers are assigned the values of the channel states , the RBF weights correspond to the a priori probability of the channel states and the RBF width is given by , with being the channel noise variance-provides the conditional density function of each legitimate -QAM symbol [16].…”

Section: Jacobian Logarithmic Rbf Equalizermentioning

confidence: 99%

“…Since then, various nonlinear structures applied to the problem of adaptive channel equalization have been investigated, such as the multilayer-perceptron network [9]- [11], polynomial-perceptron structure [12], the functional link structure [13], the self-organizing map [14], the recurrent network of [15] and RBF networks [16]. Chen, Mulgrew, McLaughlin, and Grant [6], [16], [17] promoted the employment of RBF based equalizers, which exhibit a structure identical to that of the optimal Bayesian symbol-decision based equalizer. Therefore, RBF equalizers can rely on optimal detection theory [18].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Burst-by-burst adaptive turbo-coded radial basis function-assisted decision feedback equalization

Yee

Liew

Hanzo

2001

IEEE Trans. Commun.

View full text Add to dashboard Cite

Abstract-The performance of the proposed radial basis function (RBF) assisted turbo-coded adaptive modulation scheme is characterized in a wideband channel scenario. We commence by introducing the novel concept of the Jacobian RBF equalizer, which is a reduced-complexity version of the conventional RBF equalizer. Specifically, the Jacobian logarithmic RBF equalizer generates its output in the logarithmic domain and hence it can be used to provide soft outputs for the turbo-channel decoder. We propose using the average magnitude of the log-likelihood ratio (LLR) of the bits in the received transmission burst before channel decoding as the channel quality measure for controlling the mode-switching regime of our adaptive scheme. [7]. However, there is little information concerning its potential in either BbB AQAM or in forward error correction (FEC)-coded scenarios, despite the advantageous interactions of RBF-aided DFE BbB AQAM in conjunction with turbo FEC, which were demonstrated in our preliminary work [8]. Hence in this contribution, we set out to document these interactions. IndexThe symbol-spaced channel output can be defined by (1) where ; channel impulse response (CIR); channel's input sequence; noiseless channel output.Paper approved by C. Schlegel, the Editor for Coding Theory and Techniques of the IEEE Communications Society. Manuscript received April 5, 2000; revised March 20, 2001. This work was supported by the EPSRC, the U.K., and the Commission of the European Communities, Brussels, Belgium.The authors are with the Department of Electronics and Computer Science, University of Southampton, Highfield Southampton SO17 1BJ, U.K. (e-mail: lh@ecs.soton.ac.uk).Publisher Item Identifier S 0090-6778 (01) [17] promoted the employment of RBF based equalizers, which exhibit a structure identical to that of the optimal Bayesian symbol-decision based equalizer. Therefore, RBF equalizers can rely on optimal detection theory [18]. In this contribution, we use a decision feedback assisted RBF equalizer, which involves its previous decisions for either RBF subset center selection [17] or space translation [19], [20], in order to reduce its computational complexity. Both versions of the DFE realize the same optimal solution, but the space translation based version [19], [20] requires less storage for the RBF centers, since the centers corresponding to different decision feedback vectors are equivalent in the translated space [19], [20]. Therefore the latter version is preferred in hardware implementations.BbB AQAM schemes [21] employ a higher-order modulation scheme in a certain transmission burst, when the channel quality is favorable, in order to increase the throughput and conversely, a more robust, lower-order modulation scheme is utilized in those transmission bursts, where the instantaneous channel quality drops [1]. The performance benefits of the RBF DFE [17] have been documented in the context of AQAM over dispersive wideband mobile channels in [22], [23], demonstrating that a certain target bit error rate ...

show abstract

Section: Discussionmentioning

confidence: 97%

Section: Introductionmentioning

confidence: 99%

Section: Jacobian Logarithmic Rbf Equalizermentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Burst-by-burst adaptive turbo-coded radial basis function-assisted decision feedback equalization

Yee

Liew

Hanzo

2001

IEEE Trans. Commun.

View full text Add to dashboard Cite

show abstract

“…Examples include neural network equalizers and multiuser detectors in communication systems [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15]. Typically, sample-by-sample adaptation is needed for practical applications to meet real-time computational constraints, and the training of neural network classifiers is usually done using some stochastic gradient algorithm based on the mean square error (MSE) criterion.…”

Section: Introductionmentioning

confidence: 99%

Adaptive near minimum error rate training for neural networks with application to multiuser detection in CDMA communication systems

2005

Self Cite

View full text Add to dashboard Cite

Adaptive training of neural networks is typically done using some stochastic gradient algorithm that aims to minimize the mean square error (MSE). For many classification applications, such as channel equalization and codedivision multiple-access (CDMA) multiuser detection, the goal is to minimize the error probability. For these applications, adopting the MSE criterion may lead to a poor performance. A nonlinear adaptive near minimum error rate algorithm called the nonlinear least bit error rate (NLBER) is developed for training neural networks for these kinds of applications. The proposed method is applied to downlink multiuser detection in CDMA communication systems. Simulation results show that the NLBER algorithm has a good convergence speed and a small-size radial basis function network trained by this adaptive algorithm can closely match the performance of the optimal Bayesian multiuser detector. The results also confirm that training the neural network multiuser detector using the least mean square algorithm, although generally converging well in the MSE, can produce a poor error rate performance. r

show abstract

Neural Networks and Applications to Communications

Μανωλάκος

2003

Wiley Encyclopedia of Telecommunications

View full text Add to dashboard Cite

Neural Networks are highly interconnected distributed information processing architectures that are built by connecting together many simple processing elements. They come in many different flavors but all share the ability to learn i.e. to adapt their structure to capture the process that generates information in their environment, and utilize the internally formed representation of that process to predict , or generalize , when presented with unseen novel patterns. In this article we first introduce the basics of neural networks by describing the structure of commonly used neuron processors, some popular network architectures and learning rules. Then we explain why neural network methods have been utilized so extensively in the field of communications by presenting typical snapshots of their use in a representative subset of communication applications. Our intention is to introduce the interested non‐expert reader to the subject matter and not to present a comprehensive survey of the very large related literature.

show abstract

Complex-valued radial basis function network, Part II: Application to digital communications channel equalisation

Cited by 129 publications

References 5 publications

Burst-by-burst adaptive turbo-coded radial basis function-assisted decision feedback equalization

Burst-by-burst adaptive turbo-coded radial basis function-assisted decision feedback equalization

Adaptive near minimum error rate training for neural networks with application to multiuser detection in CDMA communication systems

Neural Networks and Applications to Communications

Contact Info

Product

Resources

About